I have successfully run an LEfSe that has found ~400 significantly discriminative features in my data. I would like to plot this on a tree, but this creates an unreadable legend with too many overlapping colors.
I know that in order to decrease the numbers of significant features I can increase the LDA cutoff to be higher than the default of 2. But is there a way to create a plot with only the top 50 most significant labeled, but the other 350 nodes still colored (for each class) but without a label?
Thanks for the help.